&
Minnesota
River Valley
Counties
Minnesota County Biological Survey
Division of Ecological
Ecological Resources
Minnesota Department of Natural
Natural Resources
photo by Fred S. Harris, MNDNR
Native Plant Communities
Rare Species of the
Native Plant Communities and Rare Species
of
The Minnesota River Valley Counties
September 2007
Minnesota County Biological Survey
Division of Ecological Resources
Department of Natural Resources
Box 25, 500 Lafayette Road
St. Paul, MN 55155
Biological Report No. 89
2. Overview of the Quaternary Geological History of the Minnesota
River Watershed
Carrie E. Jennings
“The Minnesota River valley is truly the most striking
and scenic feature of all south-central Minnesota. It
is a narrow sliver of wooded hill slopes in the vast
plains to north and south, and it holds within it a
diversity of geologic features such as rugged granite
knobs on the valley floor, boulder-gravel river bars,
broad sandy terraces, gentle colluvial slopes – and a
stream along the axis that is almost tiny in the context
of these major features” (Wright 1972).
THE WATERSHED—A GLACIAL TROUGH
RUNS THROUGH IT
he Minnesota River f ows down the centerline
of the broad glacial trough formed by ice of the
Des Moines lobe, which dominates the topography
of the southern half of Minnesota (Fig. 2.1). This
trough is so wide and gently-sloping that it is
almost undetectable from the ground. The valley
of the Minnesota River , on the other hand, is an
arresting feature that was created abruptly after the
Des Moines lobe retreated and is still affecting the
evolution of the landscape today.
T
The ice lobe that formed this statewide trough was
just the last of several that entered Minnesota during
the last glaciation (and this last glaciation was just
one of several that af fected Minnesota during the
last two million years). Ice lobes were dynamic
tongues of ice that extended far south of the main
ice sheet. Rather than being the slow inexorable
movement of ice that is typically envisioned,
the ice lobes advanced and stagnated quickly
and repeatedly until they eventually drained the
slower reservoir of ice that was feeding them. The
lobes were directly controlled and fed by focused,
fast-moving zones within the ice sheet called ice
streams (Patterson 1996, 1997).
The Des Moines lobe, which extended south
from the main Laurentide ice sheet that covered
much of Canada and parts of the northern United
States during the last ice age, had more than a
dozen advances, driven by ice stream dynamics,
between approximately 40,000 and 1 1,000 years
ago (Clayton and Moran 1982, Patterson 1996).
2.1
Each advance eroded a bit more from the land
it covered and deposited it near the edge of the
lobe, creating a suite of identi f able landforms
at each ice mar gin. Geologists collectively refer
to the sediment of all of these advances of the
Des Moines lobe as the New Ulm Formation,
even though each advance may have been subtly
different in its composition. The sediment, or till,
deposited by the ice was nearly an equal mixture
of clay, silt and sand, with a predictable amount
and assortment of pebbles in the f ne-grained mud.
Till was created as overridden sediment and rock
fragments were either frozen into the glacier or
dragged along by the moving ice and eventually
conveyed towards the ice mar gin. This material
was thoroughly ground up and mixed while in the
shear zone at the bottom of the glacier.
The processes that occur beneath an ice sheet are
diff cult to study but drilling through and imaging
the base of modern ice streams in West Antarctica
have elucidated them. The boundaries of the
subglacial shear zone change with the temperature
of the bottom of the ice and these changes control
whether material is eroded or deposited, frozen in
or melted out. In areas of heat from friction, debris
is released from the ice and left behind, forming
a till sheet. Where the temperature is reduced at
the glacier bed, sediment is frozen into the ice and
carried away, resulting in erosion. The presence
or absence of water also plays an important role
in how these processes vary over time, so erosion
and sedimentation vary with the temperature and
melting point of the ice. Where the shearing force
of the ice is reduced by lubricating water, sediment
that was dragged along is left behind.
Where
subglacial water has drained away , the shearing
force increases and a thicker pile of sediment is
dragged along with the ice. Ice thickness also
comes into play directly because thicker ice
can drag along more sediment than thinner ice.
Complicated though these processes may seem,
we have evidence to show that these processes
operated in the ancient ice lobes of the Laurentide
ice sheet and effectively removed debris, mixed it
thoroughly, moved it for hundreds of miles, and
then spread a fairly uniform blanket of the end
product over the region.
remains in the Minnesota River valley is exposed
in the valley walls and forms the competent
(indurated) gray-brown to yellow bluffs along the
lower reaches of tributary streams to the Minnesota
River. More than a million years worth of glacial
advances is preserved in the banks of the Yellow
Medicine River in Upper Sioux Agency State
Park.
Beneath the till of the Des Moines lobe lie layers
of older glacial units from previous glaciations;
these layers vary in thickness and completeness
of preservation (Gilbertson 1990, Patterson et
al. 1995). The truncated stack of older tills that
Over two million years worth of sediment is
layered in the better-preserved stack of old debris
Fargo,
ND
ir
Pra
au
ote
ie C
James
Lobe
Minneapolis/
St. Paul, MN
Sioux Falls,
SD
Des Moines
Lobe
Des Moines,
IA
Figure 2.1: Location of the Des Moines and James Lobes at their maximum extents.
2.2
beneath the Coteau des Prairies (Prairie Coteau)
farther west (Fig. 2.1) (Balco et al. 2005). Near the
prow of the Coteau, along the North Dakota-South
Dakota line, drill holes penetrate 800 feet of glacial
sediment before reaching the bedrock beneath
(Beissel and Gilbertson 1987). This highland,
while dramatic, is not especially erosion resistant;
it was simply bypassed this last time around by ice
lobes on either side of it and thereby preserves a
thick stack of older glacial sediment (Gilbertson
1990, Lineburg 1993, Patterson et al. 1995).
We do not know exactly what the land looked like
before the advances of the Des Moines lobe and
previous ice advances, but expect that it originally
looked a lot like the unglaciated areas of North
Dakota, west of the Missouri River (Bluemle
1991). Earlier ice lobes from older glaciations
most likely initiated the path later followed by the
Des Moines lobe, which was the last in a long line
of ice advances from the northwest as indicated
by the similar composition of the tills. Perhaps
the early ice advances initially exploited a small
river valley. The direct evidence for this has been
removed, smeared out across Minnesota and Iowa,
and deposited as far away as the Gulf of Mexico by
continent-spanning meltwater streams. So we can
only make an educated guess based on the nature
of the unglaciated landscapes nearby.
As global climate warmed about 14,000 years ago,
bringing an end to a 100,000-year-long ice age, the
Des Moines lobe, rather than simply slipping back
into Canada, began the long, stuttering process
of advancing and stagnating in f ts and starts,
eventually leaving Minnesota about 1 1,000 years
ago. The warming climate had actually created
the conditions for this style of ice lobe advance by
lubricating the base of the ice streams and lobes.
Today, it is this record of ice stagnation that is best
preserved (Patterson 1997). Each indecisive margin
is marked by an area where debris accumulated
in a hilly band. In some cases, the stagnation of
the lobe—which had become heavily crevassed
as it wasted—was marked farther from the main
margin by faint, crisscrossing pattern of debris that
settled into the crevasses, forming low intersecting
ridges. (Kemmis et al. 1981 and references therein,
Patterson 1996). Beyond the ice, arcuate braided
streams carried meltwater around the front of the
melting ice but were not erosive enough to carve
deep valleys (Leverett 1932, Patterson et al. 1999).
Rather, they were debris-choked, shallow , and
broad. After circumscribing the lobe, they f owed
southeast down the regional slope, most likely
near the present location of the Minnesota River ,
but prior to its formation. Some of the former ,
ice-marginal stream channels are still occupied
by lazy reaches of modern rivers such as the
Redwood, Cottonwood, and Watonwan (Map 1
[inside cover], Fig. 2.1). These rivers still take the
long way around, looping around the margin of the
phantom ice lobe. In other cases, modern streams
have taken short cuts to the more recently formed
Minnesota River, with courses that are more in line
with the modern gradient. For example, theYellow
Medicine River begins in one former ice-marginal
stream channel, then cuts across the broad, level
till plain in an east-northeasterly direction until it
intersects a younger ice-mar ginal stream, thereby
taking a more direct route to the master stream, the
Minnesota.
In areas away from the low-gradient braided
streams, the ice lobe trough was not shaped to
allow water to drain readily, and pooled water, from
small puddles to lar ge lakes, formed as the lobe
melted. Most of the large lakes of glacial meltwater
eventually drained, but shallow depressions
that held water after rains persisted. The ten to
twenty feet of f ne-grained, dense, gray, clayey till
deposited in the trough by the Des Moines lobe did
not allow rainwater and snow melt to percolate into
the substrate very quickly (Patterson et al. 1999).
As seasonal water perched, wetlands formed,
covering about half of the area of the trough. Most
of these wet areas have been intentionally and
progressively drained for farmland.
After the Des Moines lobe’s f nal retreat into North
Dakota, another lobe, the Red River lobe—fed by
a different ice-stream source—took advantage
of the newly vacated lowland, advancing south
2.3
from the area we now know as Lake Winnipeg. It
created a series of broad stagnation moraines north
of Ortonville, Minnesota, collectively named for
the county they lie in, Big Stone.
Waseca, Steele, Le Sueur, Rice, Scott, Carver, and
Hennepin counties (Map 1). These eastern hills
fringed the former expanses of prairie (now corn
and soybean f elds) that occupied the f at, central
trough of the lobe; they formed effective f rebreaks
ICE MARGINAL LANDSCAPES—
against the windswept blazes that moved from
HUMMOCKS AND HILLOCKS
west to east across prairie regions of the state,
The thin ice of the punctuated advances of the Des
enabling the development of woodland and forest
Moines lobe moved a lot farther than one might have vegetation along the margins of the prairie.
predicted—splintering into a spruce forest in Des
Moines—the stumps are there to prove it (Clayton
Slightly impractical farming areas, the hummocky
and Moran 1982, Bettis et al. 1996). The advances
zones were probably the second choices of
were most likely encouraged by pressurized water
homesteaders and were more likely settled by
beneath the ice that allowed it to glide by reducing
homesick immigrants for nostalgic reasons
friction. Advances probably ended in stagnation
because they looked like a left-behind “home” in
when the water making the long glide possible
Ireland, Czechoslovakia, or central Germany. The
was siphoned of f through subglacial valleys.
hummocks and swales formed as repeated advances
These tunnel valleys formed at the ice margin and
of ice buckled the debris-covered, stagnant ice,
worked their way up into the ice lobe, draining a
which then wasted slowly and irregularly . The
large portion of it and causing it to skid to a halt
dirt and ice mix lasted a long time—long enough
(Patterson 1996). Today these valleys are marked
for trees to establish themselves in the wet debris.
by mile-wide depressions up to 10 miles long and
Ducks may have swum in the pools of water that
frequently contain long lakes (e.g., Lake Benton)
developed on the ice.The dirt on top of the stagnant
or chains of lakes (e.g., the lakes near Fairmont,
ice was wetted from below by melting ice, making
see Map 1). The lobe advanced and stagnated
it unstable and causing dirt, rocks, trees, and ducks
again and again, at least thirteen times, with each
to be rearranged for centuries on the ever-changing
advance a little less extensive than the previous
surface of the buried ice. Ultimately, all of the ice
one. This was because the lobe was progressively
melted. The hills we see today represent the f nal
drawing down the reservoir of ice.
resting place of the glacial sediment in the very
last holes in the stagnant ice surface.
Each advance also eroded the ice-lobe trough
a bit more, conveyed that material closer to the
In some areas, water associated with melting
ice margin, and piled it near ice and debris from
ice, or simply gravity , sorted the till into better the previous advance. In this way , the mar gins
organized layers of like-sized grains. Sand and
of the lobe—especially those along the east and
gravel were carried away from the ice by f owing
west sides, which received more total advances—
rivers. Silt and clay moved even farther , though
accumulated thick stacks of dirt and stagnant ice.
it also settled out in the still waters of proglacial
The slow decay of dirt-covered ice piles created
lakes formed by ice or moraine dams near the ice.
the complex hilly areas that now mark the former
Silt and sand became airborne with the persistent
limits of the ice (Clayton 1967, Patterson 1996).
and strong glacial winds. Sand particles stay lower
The western mar gins are marked by rolling hills
to the ground and saltate (bounce), continuing
that cross diagonally across Jackson, Murray ,
to dislodge particles that have settled out.
The
southwestern Lyon, Lincoln, and western Yellow
windblown silt settled out in a blanket of loess that
Medicine counties (Map 1).The eastern margins are
thickens towards the Mississippi River (Mason
marked by the oak-dotted hillocks and hummocks,
et al. 1999). Most of this sorting occurred before
specked with lakes and wetlands, in Freeborn,
vegetation became established on the landscape.
2.4
We are a state of imported dirt. Everywhere where
there is sediment in Minnesota, it is likely to be
the long-traveled deposits of an ice lobe or the
re-sorted material described above. Pick up any
rock; it was most likely imported from Manitoba
or Ontario, as was the gray clay or red sand of
the till matrix. Our rivers proceed to expose and
erode this glacial sediment and deposit it in the
nearest slackwater area, either in the streambed or
f oodplains or maybe a lake along the way . Some
sediment might make it all the way to the Gulf of
Mexico, or at least Lake Pepin.
the only outlet to the lake for much of this time
was to the south, through the Big Stone moraine.
This overf ow from glacial Lake Agassiz made
the initial incision into the newly vacated glacial
trough, starting the process of formation of what
eventually became known as the Minnesota River
valley. The early river that spilled south from the
vast, cold, iceber g-f lled waters of glacial Lake
Agassiz is referred to as glacial River Warren.
The glacial till and the reworked byproducts
of glaciation form the base for the rich soils of
Minnesota. These soils were created over the
past 10,000 years as the minerals in the glacial
sediment were broken down. The slow movement
of water through the till dissolved certain minerals
and deposited them deeper in the soil pro f le. The
mineral soil was amended with decaying organics
from the succession of plants that occupied the land
surface. Ten-thousand years worth of root activity,
frost, f re, burrowing, and acid leaching from leaf
litter have all played a role in forming the soils we
depend on in the Minnesota River watershed.
THE GREATEST LAKE
In places, meltwater from the wasting ice was
temporarily trapped—for at least a dozen years
in the basin of glacial Lake Minnesota, which
covered most of Blue Earth County in addition to
parts of other contiguous counties (Patterson and
Hobbs 1995), and for at least half a century in the
basin of glacial Lake Benson, which covered large
portions of Swift, Lac Qui Parle, and Chippewa
counties (Rittenour et al. 1998) (Fig. 2.2). As the
Red River lobe retreated, meltwater ponded behind
the Big Stone moraine for about a millennium.
The resulting glacial Lake Agassiz (Upham 1890,
1895) was one of the lar gest freshwater lakes in
the world, eventually occupying 200,000 square
miles, an area greater than all of the Great Lakes
combined, covering much of what are now western
Minnesota, eastern North Dakota, Manitoba, and
western Ontario (Fig. 2.2) (Thorleifson 1996). The
Red River lobe dammed the water to the north and
2.5
Overf owing water from glacial Lake
Agassiz
created glacial River Warren some time around
11,500 radiocarbon years before present (Clayton
and Moran 1982, Matsch 1983) and the glacial
River Warren outlet was occupied until about
10,900 radiocarbon years before present.
Two
other outlets were intermittently used by Lake
Agassiz, one to the east, through Lake Nipigon
and thence into Lake Superior , active between
10,900 to 10,300 (Thorleifson 1996), and one
to the northwest in the Fort McMurray area of
Canada through the Mackenzie River and thence
to the Arctic Ocean between 10,000 and 9,600
radiocarbon years before present (Lowell et al.
2005). When Lake Agassiz drained through these
other two outlets, the southern outlet was not used.
River Warren was probably reoccupied after 9,600
but f nally lost glacial lake dischar ge forever by
8,200 radiocarbon years before present.
The retreat of an ice sheet is not always a steady
affair, however, and so the reign of Lake Agassiz
was somewhat complicated. The lake did not
successively take lower and lower outlets that
were exposed as the ice sheet withdrew to the
north. The ice lobes, driven by the independent
ice streams, behaved erratically and shot out from
the ice front at various times, rerouting and even
blocking meltwater dischar ge. It appears that a
late readvance of ice in the Lake Superior basin
blocked the eastern outlet of LakeAgassiz through
Lake Superior for a long enough interval to allow
Lake Agassiz levels to rise to the elevation of the
southern outlet a second time. The ever-changing
part of the lake shore represented by the oscillating
ice front was not the only factor complicating the
drainage history of LakeAgassiz. The land had also
Hudson Bay
SAS
K AT
CHE
WA
N
Glacial Lake Agassiz
MANIT
OBA
O N TA R I O
NORT
H
DAKO
TA
S
ke
a
L
u
r
rio
e
p
M I N N E S O TA
SOUTH
Glacial
Lake
Benson
D A KO TA
WISCONSIN
MIN
ER
IV
NE
SOTA R
Glacial
Lake
Minnesota
Fig 2.2: Locations of glacial lakes Agassiz (after Teller et al. 1983), Benson, and Minnesota (after Hobbs and Goebel
1982) at their maximum extents. Note: these lakes were present at different times.
2.6
been rebounding gradually as the weight of the ice
sheet was lessened, more so in places where the
ice was thicker. So the northern part of the basin
was rising more than the southern part and that too
has played a part in the occupation of the various
outlets of Lake Agassiz (Thorleifson 1996).
River Warren easily cut through the stack of older
tills and saprolite (weathered bedrock) that lay in
the trough of the Des Moines lobe. In places it
reached harder rock and could not erode deeper .
It exposed old, pink quartzites near New Ulm;
very old granitic rocks on terraces near Ortonville,
Odessa, and Granite Falls; and the even older
gneisses between Redwood Falls and Morton.
These resistant rocks were local base levels that
constrained the depth of erosion of River Warren
(Matsch 1983) and required it to become wider
to accommodate the f ow. They are still the sites
of rapids or waterfalls—collectively called nick
points—as can be inferred from the names of some
of the towns.
As the river swept the clay and grit of the weathered
rock away it exposed the strange knobby bedrock
surface that lay beneath the saprolite . In a few
places, pebbles swirling in River Warren eddies
scoured potholes. But in most places the river just
removed the loose material leaving a strange but
wonderful undulating surface. This landscape is
well exposed on the route through the Big Stone
National Wildlife Refuge near Odessa.
This
chemical weathering front may have never seen
the light of day were it not for River Warren, and
presents a very odd landscape for us to explore.
The rocky outcrop area in Granite Falls stretches
for miles and is a disor ganized, lumpy moon-like
landscape (Johnson et al. 1998, Patterson et al.
1999, Patterson and Boerboom 1999). The river
stripped most of the saprolite from a bedrock
surface that was shaped by the slow chemical
weathering of hard rock in a tropical setting, with
the partially dissolved rock blanketing it during
this process. This is the way that rock turns to
mush in places like Brazil today . The undulation
of the rock surface is a result of the weathering
front proceeding unevenly , going deeper along
2.7
fractures and joints than over intact surfaces. In
some places completely isolated rounded remnants
of rock called corestones remained in the saprolite.
These remnants formed as the rock around them
dissolved. When the saprolite was swept away
by glacial River Warren it left the lar ge rounded
rocks in place (Patterson and Boerboom 1999).
The saprolite still remains in a few patches near
the surface and has been mined near Redwood
Falls and Sacred Heart for its unique clay mineral,
kaolinite. Kaolinite is the stable end-product of
long-term chemical weathering and is valued for
its purity. It is used in porcelain and coated color
paper, among other things.
There are waterfalls or rapids (nick points) on all
of the tributary streams to the modern Minnesota
River. The tributaries instantly developed
waterfalls when glacial River Warren was incised
below the levels of its tributaries. These waterfalls
gradually moved upstream or became a series of
rapids or nick points. In this way the lower reaches
of all of the tributaries to glacial River Warren are
adjusting their gradients to meet the new level of
River Warren, a journey they are still on today
even though River Warren ceased to f ow millennia
ago. Even now, only the lower 5–10 miles of each
major stream is adjusted to the elevation of the
valley f oor of the Minnesota River . Look again
at the location of the waterfalls, towns or dams on
these tributary rivers. Their names and locations
reveal the progress of nick point migration during
the last 10,000 years.
The process does not stop just because we live here
now. The nick points will continue to move up the
tributaries and the tributaries to the tributaries, and
so on, until the entire watershed of the Minnesota
River valley has adjusted. Given the progress over
the last 10,000 years, one might expect another
glaciation to occur before the result is achieved.
However, changes we have made to the drainage
system have accelerated the rate at which this
process would typically take place.
Upriver from Mankato, glacial River Warren was
primarily a down-cutting stream (Johnson et al.
1998 and references therein). This means that
terraces there are controlled by the depth of
erosion of the river, usually to a level of resistant
rock, and that surfaces along the valley are not
related to stable lake levels and beaches of Lake
Agassiz (Johnson et al. 1998). Some are channels
that were brie f y occupied before the erosive
water became focused in one main channel—for
example, Watson Sag (map 1) (Kehew and Lord
1986, Patterson et al. 1999). So depositional sand
and gravel terraces are not a common feature of
glacial River Warren in its upstream reaches.
Where one does see sand or gravel f anking the
upper Minnesota River valley , it is most likely
remnants of braided glacial outwash streams
or deltas that had formed in the wide, slower
parts of the River Warren. There is a broad and
complicated area of deltaic and outwash sand in
the area near Appleton. It predates the formation
of the glacial River Warren trench and is linked
to glacial Lake Benson, which formed during ice
lobe retreat (Patterson et al. 1999).
and the lake must have suddenly and completely
drained through a valley trending north. We see
no direct evidence of that valley now; it has been
completely obscured by the later drainage events
of glacial River Warren. However, we do see the
evidence for a proglacial lake, the outlets for which
were progressively moving north towards lower
ground. Some things can never be de f nitively
proven, but other explanations for the bend in the
river, such as a change in bedrock near the area of
the bend (Wright 1972), don’t seem to explain the
existence of the bend as comprehensively.
A right-angle bend exists in the Minnesota River
valley at Mankato. It is likely that this bend was
inherited from the course of an earlier stream that
developed while the Des Moines lobe was in retreat.
This stream appears to have started as an icemarginal stream that was then used to drain glacial
Lake Minnesota. There is no really good reason
for a river to make a sharp left-hand turn otherwise
and river courses tend to get reused. Glacial Lake
Minnesota formed when the Des Moines lobe
margin lay in North Mankato.The lake spread south
almost to the Iowa border and deposited a fairly
continuous blanket of silt and clay that buried the
till in much of the Blue Earth watershed. The lake
was initially deep enough to spill south through
Union Slough into Iowa. Then a lower outlet to
the east opened and the lake drained through what
are now Waseca and Steele counties. The channel
occupied by the chain of lakes that starts in Le
Sueur County and trends northeast through Rice
County and includes Cannon and Sakatah lakes
is another presumed outlet for this lake (Patterson
and Hobbs 1995). When the ice lobe receded from
North Mankato, even lower ground was revealed
2.8
Whatever the cause, the bend in the river at
Mankato must have set up a mighty eddy. No civil
engineer would deliberately create a channel of that
shape for fear of erosion along the valley wall and
the f otsam and jetsam that would accumulate in a
giant, backward circling eddy. The effort of getting
the water through here might have even created a
hydraulic dam, backing water up in the channel
as the river worked its way through the extreme
bend, perhaps f ooding the newly formed mouth
of the Blue Earth River as evidenced by deposits
of sand and gravel that are currently being mined
several miles up from the mouth of the river.
The river encountered even more engineering
problems downstream of Mankato: buried valleys
that had been f lled by deposits of earlier glacial
episodes. The valleys were relatively easy to ream
out and reoccupy but the river course was also
interspersed with resistant, bedrock-reinforced
areas. These sedimentary rock layers of sandstone
and limestone thus created local nick points that
would have migrated up a river (Johnson et al.
1998) already complicated by large f ows of water
and strange hydraulics. Some of this history is
recorded in the walls of a sand and gravel mine
on the broad terrace near Kasota. Fifty-foot high
angled beds of sand—foresets—are evidence for a
very deep, pooled part of the river valley thatf lled
with sand as a bar or delta migrated across tof ll it.
The river lies far below this terrace now so it was
eventually able to cut through the rock that is also
mined here (Kasota stone is the local name given
the rock) and form a narrower channel.
With the cessation of lake water dischar ge from
glacial Lake Agassiz, the River Warren valley
became grossly oversized for the regional
precipitation brought to it by its tributaries.
Tributaries to the Minnesota River , the successor
to glacial River Warren, were in the process of
adjustment to the level of the deeply incised valley
and therefore were carrying as much sediment as
they could handle, which they delivered to the
valley. The Minnesota River was not ef fective at
carrying this sediment away and the tributaries
built fans at their mouths. Lakes formed behind
these fans and long reaches of the Minnesota River
were dammed (Zumber ge 1952) to create river
lakes that include Big Stone Lake (at the fan of
the Whetstone River), Lac Qui Parle Lake (at the
fan of the Lac Qui Parle River), and Marsh Lake
(at the fan of the Pomme de Terre River) (Wright
et al 1998). Much farther south, but in a similar
fashion, Lake Pepin was created by a tributary fan
from the Chippewa River of Wisconsin (Wright
1990, Wright et al. 1998). Water in Lake Pepin was
initially backed up all the way to downtown St.
Paul (Zumberge 1952, Wright et al. 1998).
River lakes age in the same way that an artif cially
dammed river does: they f ll with sediment carried
by the river. All of the river lakes in the state are
gradually f lling in as sediment is delivered to
these shallow pools. The Minnesota River valley
is f lling in, or aggrading, everywhere; it is just
more obvious in these river lakes. It is also more
obvious to us now because modif cations we have
made to the landscape and drainage network have
sped up the rate of sediment delivery (W right et
al. 1998).
Landscapes that avoided the most recent
glaciation, but were glaciated previously (Fig.
2.1, southeastern and southwestern regions of
Minnesota), are veined with complexly branching
tributaries of river systems. Natural lakes simply
do not exist in these corners of the state. Here,
rivers have effectively drained any formerly f at,
wetland- and lake-dotted areas in the hundreds of
thousands of years they’ve had to get the job done.
The development of this kind of dendritic drainage
2.9
Bedrock of the
Minnesota River Valley
Fred S. Harris
T
he knobs of ancient granite-like bedrock
exposed in the Minnesota River valley
upstream from New Ulm include some of the
oldest rocks discovered at the earth’ s surface.
Several geologists estimate that many of these
“crystalline” rocks f rst formed as igneous rocks
from molten magma that cooled very slowly
deep below the earth’ s surface as long as 3.6
billion years ago (Grant 1972), when the core
of North America was being formed. Once
formed, these early rocks underwent extreme
heat and pressure over the next 1 to 1.5 billion
years, which altered their crystalline structure
and transformed them into metamorphic rock.
These rocks include many variants of gneisses
and diorites that may be found today in several
parts of the valley, such as the Morton Gneiss,
which is exposed in the town of Morton.
Other crystalline bedrocks exposed in the valley
are far younger in age. These include igneous
rocks, such as granite and gabbro, which f owed
into (or “intruded”) cracks in the older gneisses. Examples include the Sacred Heart Granite,
which formed about 2.6 billion years ago, and the
Cedar Mountain Complex near Franklin, which
dates to 1.8 billion years ago (Grant 1972). Cedar Mountain is a classic example of an igneous
intrusion, in which magma f owed into cracks
in older rocks. This complex includes an outer
ring of gray rock (the Cedar Mountain Gabbro),
which encircles an inner zone of pinkish crystalline rock (Cedar Mountain Granodiorite) (Bury
1958, Grant 1972). Today, the Cedar Mountain
complex forms the highest point of relief within
the Minnesota River valley (Schwartz and Theil
1954).
Sioux Quartzite is a younger bedrock layer exposed in the Minnesota River valley only near
New Ulm. Many more exposures of it are located farther south and west. This is a pink-to-purplish rock that formed from sandstone deposited
in a shallow sea (Ojakangas and Matsch 1982).
The sandstone (composed of quartz grains)
was then transformed into a metamorphic rock
(quartzite) by heat and pressure, possibly due to
severe compression from tectonic plate movement, which cemented the sand grains into much
harder rock. Its dark, purplish color comes from
iron oxide coatings on the quartz grains (Austin
1972).
In the eastern part of the Minnesota River valley, sedimentary rocks are exposed that originated as sediments in shallow seas that occupied much of the North American interior 500
million to 430 million years ago (Ojakangas
and Matsch 1982). These marine environments
contained a rich diversity and abundance of
plants and animals, which remain embedded
in some of the rocks as fossils. The Canadian
Shield, which underlies northern, central, and
western Minnesota, was not inundated by these
shallow seas; thus these sedimentary rocks are
absent from the western part of the Minnesota
valley (Ojakangas and Matsch 1982). Major
sedimentary rock units exposed in the lower
Minnesota valley in order from youngest (400
million years old) to oldest (over 500 million
years old) include the following:
•
•
•
•
Platteville Limestone: a 35-foot thick layer
of hard, fossil-rich limestone on top of the
St Peter Sandstone. This is visible along the
upper edges of the steep rock cliffs along the
Mississippi River valley in the Twin Cities.
St. Peter Sandstone: a 155-foot thick layer
of mostly white sandstone that makes up
most of the exposed cliffs along the Mississippi River in the Twin Cities.
Prairie du Chien Group (Shakopee and
Oneota Dolomite): a 410-foot thick layer
mostly composed of dolomitic limestone.
Oneota Dolomite is the locally known Kasota Stone being mined from a glacial river
terrace near Kasota, as described above.
Jordan Sandstone: a 1 15-foot thick layer
of whitish to yellowish sandstone exposed
near Mankato and Jordan.
network is something that is just getting started
in the recently deglaciated landscape of the
Des Moines lobe. Eventually , the Des Moines
lobe trough should be drained by a dendritic
network every bit as complex as what we see
in the southeastern and southwestern parts of
the state. We can see it happening. The rate and
style of drainage development, however , have
been altered by the creation of new , artif cial
tributaries such as farm tiles and ditches.
Geologists are trying to understand the complex
responses of the Minnesota River, its river lakes,
and its tributaries to these changes.
REFERENCES
Austin, G.S. 1972. The Sioux Quartzite, southwestern Minnesota. In Geology of Minnesota:
A centennial volume, ed. P.K. Sims and G.B.
Morey, 450–55. St. Paul: Minnesota Geological
Survey, University of Minnesota.
Balco, G., J.O.H. Stone, and C. Jennings. 2005.
Dating Plio-Pleistocene glacial sediments using
the cosmic-ray-produced radionuclides 10Be
and 26Al. American Journal of Science 305:1–
41.
Beissel, D.R., and J.P. Gilbertson. 1987. Geology. Part 1 of Geology and water resources of
Deuel and Hamlin counties, South Dakota. Bulletin 27. Vermillion: South Dakota Geological
Survey.
Bettis, E.A., D.J. Quade, and T.J. Kemmis.
1996. Hogs, bogs and logs: Quaternary deposits and environmental geology of the Des
Moines lobe. Guidebook Series No. 18. Iowa
City: Iowa Geological Survey.
Bluemle, J.P. 1991. The face of North Dakota.
Educational Series 21. Bismarck: North Dakota
Geological Survey.
Bury, C.A. 1958. The geology of the Cedar
Mountain Complex, Minnesota River Valley.
M.S. thesis, University of Minnesota, Minneapolis.
2.10
Clayton, L. 1967. Stagnant-glacier features of
the Missouri Coteau in North Dakota.In Glacial
geology of the Missouri Coteau and adjacent
areas, ed. L. Clayton and T.F. Freers, 25–46.
Miscellaneous Series 30 (Guidebook from the
18th Midwest Friends of the Pleistocene Field
Conference). Bismarck: North Dakota Geological Survey.
Lineburg, J.M. 1993. Sedimentology and stratigraphy of Pre-W isconsin drifts, Coteau des
Prairies, eastern South Dakota. M.S. thesis,
University of Minnesota-Duluth.
Clayton, L., and S.R. Moran. 1982. Chronology of late-W isconsinan glaciation in middle
North America. Quaternary Science Reviews
1:55–82.
Mason, J.A., E.A. Nater , C.W. Zanner, J.C.
Bell. 1999. A new model of topographic effects
on the distribution of loess, Geomorphology
28:223–236.
Gilbertson, J.P. 1990. Quaternary geology
along the eastern f ank of the Coteau des Prairies, Grant County. South Dakota. M.S. thesis,
University of Minnesota-Duluth.
Ojakangas, R.W., and C.L. Matsch. 1982. Minnesota’s Geology, Minneapolis: University of
Minnesota Press.
Grant, J.A. 1972. Minnesota River valley,southwestern Minnesota. In Geology of Minnesota:
A centennial volume, ed. P.K. Sims and G.B.
Morey, 177–96. St. Paul: Minnesota Geological
Survey, University of Minnesota.
Hobbs, H.C., and J.E. Goebel. 1982. Geologic
map of Minnesota, Quaternary geology [map].
1:500,000. Map S-1. St. Paul: Minnesota Geological Survey, University of Minnesota.
Lowell, T.V., T.G. Fisher, and G.C. Comer. 2005.
Testing the Lake Agassiz meltwater trigger for
the Younger Dryas. Eos 86 (40):365–73.
Patterson, C.J. 1997. Southern Laurentide ice
lobes were created by ice streams: Des Moines
lobe in Minnesota, U.S.A. Sedimentary Geology 111:249–61.
Patterson, C.J., and H.C. Hobbs. 1995. Surficial
geology [map]. 1:100,000. In Geologic atlas of
Rice County, ed. H.C. Hobbs, plate 3 . County
Atlas Series C-9, Part A. St. Paul: Minnesota
Geological Survey, University of Minnesota.
Patterson, C.J., and T.J. Boerboom. 1999. The
signif cance of pre-existing deeply weathered
crystalline rock in interpreting the ef fects of
glaciation in the Minnesota River valley, U.S.A.
Annals of Glaciology 28:53–58.
Kehew, A.E., and M.L. Lord. 1986. Origin and
large-scale erosional features of glacial-lake
spillways in the northern Great Plains. Geological Society of America Bulletin 97:162–77.
Kemmis, T.J., G.R. Hallberg, and A.J. Lutenegger. 1981. Depositional environments of glacial
sediments and landforms on the Des Moines
lobe, Iowa. Guidebook Series 6. Iowa City:
Iowa Geological Survey.
Patterson, C.J., A.R. Knaeble, S.E. Gran, and
S.J. Phippen. 1999. Surficial geology [map].
1:200,000. In Quaternary geology–upper Minnesota River basin, ed. C.J. Patterson, plate 1.
Regional Hydrogeologic Assessment Series
RHA-4, Part A. St. Paul: Minnesota Geological
Survey, University of Minnesota.
Leverett, F. 1932. Quaternary geology of Minnesota and parts of adjacent states. With contributions from F .W. Sardeson. United States
Geological Survey Professional Paper 161.
Washington, DC: United States Government
Printing Off ce.
2.11
Patterson, C.J., B.L. Lusardi, D.R. Setterholm,
and A.R. Knaeble. 1995. Quaternary stratigraphy [map]. 1:200,000. In Quaternary geology–
southwestern Minnesota, ed. D.R. Setterholm,
plate 2. Regional Hydrogeologic Assessment
Series RHA-2, Part A. St. Paul: Minnesota Geological Survey, University of Minnesota.
Rittenour, T.M., K.L. Geiger, and J.F.P. Cotter.
1998. Glacial Lake Benson, west-central Minnesota. In Contributions to Quaternary studies in Minnesota, ed. C.J. Patterson and H.E.
Wright, Jr., 97–102. Report of Investigations
49. St. Paul: Minnesota Geological Survey ,
University of Minnesota.
Schwartz, G.M., and G.A. Thiel. 1954. Minnesota’s rocks and waters: A geological story.
Minneapolis: University of Minnesota Press.
Teller, J.T., L.H. Thorleifson, L.A. Dredge, H.
C. Hobbs, and B.T. Schreiner. 1983. Maximum
extent and major features of Lake Agassiz. In
Glacial Lake Agassiz, ed. J.T. Teller and L.
Clayton, 43–45. Special Paper 26. St. John’ s,
Newfoundland: Geological Association of Canada.
Upham, W. 1890. Report of exploration of the
glacial Lake Agassiz in Manitoba. Geological
Survey of Canada, Annual Report 1888–89,
Part E. Ottawa: Natural Resources Canada.
Upham, W. 1895. The Glacial Lake Agassiz.
United States Geological Survey , Monograph
25. Washington, DC: Government Printing Off ce.
Wright, H.E., Jr. 1972. Physiography of Minnesota. In Geology of Minnesota: A centennial
volume, ed. P.K. Sims and G.B. Morey, 561–78.
St. Paul: Minnesota Geological Survey, University of Minnesota.
Wright, H.E. Jr. 1990. Geologic history of Minnesota rivers. Educational Series 7. St. Paul:
Minnesota Geological Survey , University of
Minnesota.
Wright, H.E., Jr ., K. Lease, and S. Johnson.
1998. Glacial River Warren and the environmental history of southeastern Minnesota. In
Contributions to Quaternary studies in Minnesota, ed. C.J. Patterson and H.E. Wright, Jr.,
131–40. Report of Investigations 49. St. Paul:
Minnesota Geological Survey , University of
Minnesota.
Zumberge, J.H. 1952. The lakes of Minnesota:
their origin and classification. Minnesota Geological Survey, Bulletin 35. Minneapolis: University of Minnesota Press.
2.12